Permissible collateral, access to finance, and loan contracts: Evidence from a natural experiment Bing Xu Universidad Carlos III de Madrid

Size: px

Start display at page:

Download "Permissible collateral, access to finance, and loan contracts: Evidence from a natural experiment Bing Xu Universidad Carlos III de Madrid"

Eric Hodge
5 years ago
Views:

1 Permissible collateral, access to finance, and loan contracts: Evidence from a natural experiment Bing Xu Universidad Carlos III de Madrid BOFIT, 2016, HELSINKI

2 Introduction Lack of sufficient collateral is a key obstacle to external finance across countries. (Hart and Moore, 1994; Stiglitz and Weiss, 1982) Inadequate legal infrastructure excludes movable assets from permissible collateral class (Campello and Larrian, 2015), while around the world, movable assets comprise half of firms total assets (Alvaerz de la Campa, 2011) Reforms in collateral law generally promotes access to external finance Campello and Larrian (2015): Romania and Eastern European countries. Aretz et al., (2015): Napolenoic Code in France. Love et al., (2015): Collateral registries promote access to finance across countries. Cerqueiro et al., (2015): Swedish collateral reform. Excessive reliance on collateral reduces bank s incentives for adequate screening, consequently allowing riskier borrowers to obtain loans and worsening credit allocation efficiency (Manove et al., 2001; Zazzaro, 2005; Jappelli et al., 2005; Assuncao et al., 2014).

3 Introduction China completed Property Law reform in 2007, allowing movable assets as permissible collateral for the first time, and bringing secured transactions more in line to international best practices. Doing Business Legal Rights Index jumps from 4 to 6 following the reform. Referred to as China s next revolution by The Economist. Few academic researches have investigated the impact of the Chinese Property Law (e.g. Berkowitz et al., 2013). Questions: Does the Property Law reform change firms debt capacity and maturity structure? Does the reform change firms reliance on informal financing channel (trade credit)? Does the reform improve credit allocation efficiency? What are the real side implications of the reform, such as firms asset mix and profitability? How does the reform shape loan contracts?

4 Institutional background: The 1995 Security Law Permissible assets: Fixed assets Equipment and motor vehicles: allowed as non-possessory security interests Inventory is in practice accepted as possessory security interests. Non-permissible assets: Account receivables Future acquired properties Any property that can t be fixed in type, quantity, and location Secured interest registration Secured interests must be registered to be enforceable No centralized registration system exists Collateral needs to be appraised and certified No clear rule on determination of priority among competing claims

5 Institutional background: The 2007 Property Law Additional permissible assets: Account receivables, including but not limited to the following Claims from sale, lease, rendering services, granting loans or other credit. Rights to charge fees from immovable property such as toll roads, bridges, tunnels, ferries, etc. Existing and future production equipments Raw materials and semi-finished goods Inventories Simplified registration process and centralized registration system General movable assets: local subset of State Administration of Industry and Commerce; require only basic information such as parties, debt, and the security Account receivables: Credit Reference Centre of People s Bank of China. Publicizing registration information of the pledge of account receivables Clear rule on determination of priority among competing claims

6 Institutional background: Secured transactions snapshot Before 2007 Secured transactions strongly favor real property as security. World Bank Group (2007):less than 7% of loans in China is secured purely by movables assets, which were mostly inventories and equipments. After 2007 During , the number of loans backed with movable assets increased by 21% per year, and the value of loans grew by 24% per year. More than 1.7 millions receivable-backed transactions have been recorded by the end of July, 2015, marking a remarkable annual growth of 51%. The registered receivable-backed transactions amount to 57 trillion RMB Source: IFC Secured Transactions Advisory Project in China (2011)

7 Data and key variables Data source: Wind Information Sample coverage: Listed firms in Shanghai Stock Exchange and Shenzhen Stock Exchange, Non-financial firms that have annual reports both before and after ,000 firm-year observations from 1,800 firms. Movratio=(Inventory+Account Receivable+Other Receivable)/Asset

8 Identification strategy Enactment of the Property Law in 2007 as exogenous shock Timing of the law is unpredicted Berkowitz, Lin, and Ma (2013): the passage of the Property Law was a surprise to the stock market. Difference-in-differences framework: The Property Law pertains to movable assets only=>firms with ex-ante high reliance on movable assets are affected more by the law change. Highmov i =1 if pre-reform median movable ratio is higher than industry median (Treated firms) Highmov i =0 if pre-reform median movable ratio is below or equal to industry median (Control firms) X includes Tangibility it-1, Liquidity it-1, Profitability it-1, Sale it-1, Age it-1, List it, Split it, and State it Firm size:100 dummies for each percentile of the size distribution (Gopalan et al., 2015) Firm FE, Year FE, Industry-Year FE, Province-Year FE β: pre-post difference in Y ijt of firms with high movable ratio, relative to the prepost difference of firms with low movable ratio Standard errors clustered at firm level (Bertrand et al., 2004)

9 Results: Debt and debt maturity 1. Debt t /Asset t-1 increases in relative terms for treated firms by 2% more, representing 6.48% of sample average. 2. LongDebt t /Asset t-1 increases in relative terms for treated firms by 2.8% more, representing 27.2% of the sample mean. 3. ShortDebt t /Asset t-1 does not show differential changes for treatment and control group

10 Results: Account payable Trade credit as substitute to formal credit (Allen et al., 2005) Implication: credit expansion could reduce firms reliance on trade credit Findings: High movable firms reduces account payable disproportionally more Effect stronger among high movable firms that expanded debt capacity after the reform.

11 Results: Who benefits the most? Theories: Collateral is usually associated with low quality borrowers (e.g. Berger and Udell, 1990, 1995; Jimenez, Salas, and Saurina, 2006). Lazy bank: providing collateral may reduce banks incentive to screen borrowers adequately (Manove, Padilla and Pagano, 2001) Implications: The debt expansion effect of Property Law should be more pronounced among low quality firms. Divide the firms into subsamples according to ex-ante firm quality proxies, and test if coefficients on Highmov i *After t are statistically equal in each subsample pairs Firm quality proxies: financially constraints (Hadlock and Pierce Index); indebtedness (Liability Ratio); Altman s Z-score; return on sale (ROS); net profitability (Netpft); and return on assets (ROA).

12 Results: Who benefits the most? Financially constrained, highly indebted, prone to bankruptcy, and less profitable firms expand the most debt capacity Potential credit allocation inefficiency (Mian and Sufi,2009, 2010;Assuncao et al., 2014)

13 Results: Asset structure Firms match debt maturity and asset maturity (Myers, 1977; Milbradt and Oehmke, 2014). Credit multiplier effects (Bernanke et al., 2000; Campello and Hackbarth, 2012). Findings: High movable firms increase disproportionally more in fixed assets investments; Credit expansion promotes fixed assets investments.

14 Results: Post-reform performance Does the reform-induced debt expansion improve firm performance? Firms expanded debt capacity after reform: Lev change Lev_inc: indicator variable equals one if the average yearly changes in debt to asset ratio for the pre-reform period is lower than that of the post-reform period Switcher: indicator variable equals one if firm s average debt to asset ratio moves from the 25% percentile in the pre-reform period, to 75% percentile in post-reform period. β 1 >0 if debt expansion improves performance Performance measures: profitability (Netprofits t /TA), EBIT/Sale, profit growth ((Netprofits t -Netprofits t-1 )/ Netprofits t-1 ), and Distress (equals one if firm s interest coverage ratio is less than one).

15 Results: Post-reform performance Performance does not improve for high movable firms that expanded debt capacity.

16 Results: Pre-trends, reverse causality and dynamic effects Validate assumptions of DID framework Trends in the outcome variables must be similar before the reform Property Law must be exogenous to firms financial decisions (Berkowitz, Lin, and Ma, 2013) Dynamics of the outcome variables after the legal reform: lead-lag model. D τ =1 in the τ th year before (after) the Property Law reform and 0 otherwise. The coefficients β τ compare the level of the dependent variable τ years before (after) the Property Law, to the year immediately before the enactment of the Property Law.

17 Results: Pre-trends, reverse causality and dynamic effects

18 Robustness tests First set of tests: other aspects of Property Law Creditor rights protection Property rights protection Second set of tests: contemporary reforms that may have differential impacts on firms with varying level of movable assets. 2005/2006: Related party transaction (Tunneling) 2005/2007: Split-share reform 2008: Unification of corporate tax Third set of tests: macroeconomics shocks that may have differential impacts on firms with varying level of movable assets. Credit condition Stimulus package Global financial crisis Collateral value

19 Robustness tests Fourth set of tests: differences in observables other than movable assets. Matched sample: treatment and control firms are as similar as possible, except movable ratio. Controls explicitly the differential reactions to Property Law as function of pre-reform firm characteristics Z: Tangibility, Liquidity, Age, Profitability, Sale, and Leverage. Fifth set of tests: variations in the exposure to treatment Pledgibility of difference classes of movable assets Availability of movable assets: Sixth set of tests: alternative definitions and samples Alternative definition of treatment and control group Alternative definition of movable assets Balanced sample Listed throughout sample period Firms never changed ownership

20 Loan contracts Question 1: Do banks issue loans at better terms to high movable firms after the reform? Question 2: Do pledging collateral after the reform improves other loan contract terms? Data: Hand-collected loan level data from firms listed in Shenzhen Stock Exchange , in total more than 10,000 loans Remove firms that borrowed only before or only after the reform=> left with more than 4,000 loans Information on: Loan size, interest rate, maturity and collateral. Dates of loan origination, borrower characteristics. Name bank, name office.

21 Results: Loan contracts Lending spread of high movable firms decreases by 46 basis points more. (15% of average). Maturity increases for high movable firms by 2 months more. Likelihood of pledging collateral increases for high movable firms by 15%. Lending spread for secured long-term loans decreases by 33 basis points more than unsecured long-term loans Secured loans become larger, but mainly for long-term loans.

22 Conclusion Debt capacity and maturity: Reform expands debt capacity and prolongs debt maturity. Informal financing channel: The reform induced debt expansion reduces firms reliance on informal credit such as account payables. Credit allocation: Reform allows financially constrained, highly indebted, prone to bankruptcy, and less profitable firms to take on most extra debt=> credit allocation inefficiency. Real-side effects: Debt expansion leads to more investment in fixed asset. Expanding debt capacity does not improve high movable firms performance. Loan contracts: Lending rates drop, loan maturity increases, and likelihood of pledging collateral increases for high movable firms. Secured loans become larger and cheaper.

23 Policy implications Collateral improves debt capacity, but it may also decrease banks screening incentives and allow riskier borrowers to take extra credit, and consequently, increases systematic risks (e.g.manove et al., 2001; Zazzaro, 2005; Jappelli et al., 2005). Credit expansion may have also caused over-leverage or overcapacity, which could lower profitability for firms in those industries and damage overall economic balancing. Highlight the importance of accompanying collateral law reform with other structural reforms to improve overall efficiency of credit allocation.

PERMISSIBLE COLLATERAL AND ACCESS TO FINANCE: EVIDENCE FROM A QUASI-NATURAL EXPERIMENT. Bing Xu. Documentos de Trabajo N.º 1750

PERMISSIBLE COLLATERAL AND ACCESS TO FINANCE: EVIDENCE FROM A QUASI-NATURAL EXPERIMENT 2017 Bing Xu Documentos de Trabajo N.º 1750 PERMISSIBLE COLLATERAL AND ACCESS TO FINANCE: EVIDENCE FROM A QUASI-NATURAL